No abstract

Research was conducted to define appropriate compound loading conditions and energy parameters required to properly control and analyze fatigue crack propagation experiments for tire sidewall applications. The effects of strain level, pulse frequency, overall cycle frequency, sample thickness, and oven temperature were screened, and strain level was shown to be the dominant variable in the region of interest. Designed experiments further confirmed that frequency (i.e., strain rate) effects upon strain energy are small at normal rates of tire deformation (equivalent to 40 Hz). However, at typical laboratory test frequencies (≤5 Hz), strain rate effects on strain energy are large, and the differences vs. results under tire conditions depend heavily on polymer type as well as test temperature. Thus, the use of strain level, strain rate, and temperature conditions which simulate the tire service environment are critical to give representative results in laboratory testing. A constitutive equation was defined which provides an excellent model for strain energy in pure (or simple) shear as a function of the principal extension ratio (i.e., strain level) at constant frequency. Therefore, computer modeling of such experiments appears straightforward using an on-line minicomputer. Fatigue crack propagation studies showed major effects of pure-shear sample thickness, processing prior to molding, different types of reference compounds, and different polymer types. Halobutyl compounds and halobutyl/EPDM/NR blends were shown to provide superior FCP resistance at a given strain or strain energy level. These results were consistent with earlier tire and laboratory data.

It is important to develop laboratory fatigue methods which are predictive of tire performance to minimize development time and costs. A basic premise of this paper is that fatigue testing based on fracture mechanics principles has major advantages over traditional methods or those which attempt to simulate the geometry of the composite tire in the laboratory. Basic fracture-mechanics theory will be reviewed, followed by summaries of previously published studies. Three new studies will illustrate the wide applicability of the methodology.

The use of internal hyperelastic solids for modeling viscoelastic deformations of rubber vulcanizates is reviewed. The model is applied in one dimension to viscoelastic uniaxial tension and uniaxial shear experiments. Step-strain relaxation tests are used to determine the model's parameters. A hyperelastic energy function, which represents the sum of the internal solids' energy functions, is obtained by least squares fitting a constrained third-order invariant expansion of the Rivlin function to the difference between the step-strain stresses and the relaxed stresses (the standard hyperelastic solid's stresses). The difference energy function is split into two parts and relaxation parameters (related to the rate of change of the internal solids' reference lengths) are selected so that numerically simulated step-strain relaxation stresses approximate the experimental values (at approximately 50 ms). The model is then used to predict the experimental results from a different type of test, cyclic strain data, at three different strain rates (cyclic frequencies). Increased stress due to increased strain rate was indicated by the model for large strains.